Recently, an acceleration sensor is in great demand, for example, in the fields of transportation such as the automobile, train, space and aeronautic industry, medical care, and industrial instrumentation. A mechanical sensor has been used conventionally, but this type is increasingly being replaced by a curved gage type using a semiconductor, or a capacitance type in order to achieve a compact size, high performance, low price, and high reliability. A semiconductor type acceleration sensor is manufactured by means of semiconductor techniques and micromachining techniques. A curved gage type acceleration sensor is configured as in the example shown in FIG. 14 (H.V. Allen et al., Sensors and Actuators, 20 (1989), pp. 153-161). In FIG. 14, reference numeral 71 represents a curved gage type acceleration sensor element; 22 represents a cantilever part; 23 represents a gage resistance part (piezoresistance element); 24 represents a connection electrode; 25 represents an upper stopper wafer; 26 represents a sensing wafer; 27 represents a lower stopper wafer; 28 represents an air gap; and 29 represents a dead weight part. When an acceleration is imposed on the dead weight part, the cantilever part is curved or distorted, and according to piezoresistance effects, the amount of electric resistance of the diffusion layer (gage resistance part: piezoresistance element) formed on top of the cantilever part changes. By means of a Wheatstone bridge circuit comprising four pieces of piezoresistance elements, an acceleration signal can be detected in the form of voltage output.
Furthermore, a capacitance type acceleration sensor is configured as in the example shown in FIG. 15 (H. Seidel et al., Sensors and Actuators, A21-A23, (1990", pp. 312-315). In FIG. 15, reference numeral 81 represents a capacitance type acceleration sensor element; 32 represents an upper counter electrode; 33 a dead weight electrode (movable electrode); 34 represents a lower counter electrode; 35 represents an upper glass; 36 represents a silicon resin; 37 represents a lower glass; 38 represents an air gap; and 39 represents a dead weight part. As for the electrodes for forming capacitance, one of the electrodes serving as the dead weight electrode 33 is disposed on top of the dead weight part 31. By detecting the amount of capacitance change accompanied by a change in distance between the fixed upper counter electrode part 32 and the lower counter electrode 34, an acceleration signal can be obtained. The capacitance type shows a great capacity change by the imposed acceleration, so that a highly accurate measurement is possible by devising a detecting circuit.
The acceleration sensor using a semiconductor is characterized in that peripheral circuits such as an amp etc. can be integrated as well. Therefore, it is possible to attain a small size and high performance by reducing the number of components and connecting points.
On the other hand, a pyroelectric infrared sensor is a thermal type infrared sensor which applies a dielectric thin film. This sensor can be operated at room temperature and has a small wavelength dependency of sensitivity. This sensor is highly sensitive among thermal type sensors. This pyroelectric infrared sensor makes use of titanate lead lanthanum (hereinafter abbreviated as PLT) as a dielectric material possessing large pyroelectric characteristics, and is usually manufactured by using a PLT film crystal-oriented to a "c-axis" which is a crystal orientation having the highest pyroelectric coefficient. In order to convert efficiently an intercepted infrared ray to a sensor output, it is necessary for the pyroelectric infrared sensor to have a PLT film which is sensitive to a thermal change corresponding to a change in the intercepted infrared amount. Therefore, a holding structure supporting the PLT film is designed, with respect to form and material, to be small in heat capacity and also such that thermal loss due to heat conduction is small.
For example, when a pyroelectric infrared sensor is used for an infrared point sensor, a conventional pyroelectric infrared sensor element comprising the central part is configured as shown in FIG. 28. Namely, a PLT film 204 having a thickness of about 3 .mu.m is disposed on the both surfaces with a lower lead electrode 223 and a upper drawer electrode 225. In order to reduce heat capacity and heat conduction, only polyimide resin films 222 and 232 hold the combination of the PLT film 204, the lower lead electrode 223, and the upper drawer electrode 225. Furthermore, the polyimide resin films 222 and 232 are held at the circumference by a ceramic substrate 207 made of ceramics which is disposed with a through-hole in the central part having a rectangular cross-section. Here, reference numerals 206, 216, and 230 represent conductive adhesives; and 208 and 209 represent connection electrodes.
The pyroelectric infrared sensor element as configured above was manufactured by a conventional method shown in FIG. 29 (a) to (f) (e.g., cf. Ryoichi Takayama, et al., "Pyroelectric infrared picture image sensor", National Technical Report, Vol. 39 (No. 4) (1993), pp. 122-133).
First, a MgO monocrystal substrate 221 having a cleavage plane of (100) and having been mirror polished is prepared. While the MgO monocrystal substrate 221 is kept at a heating temperature of 600.degree. C., a ceramic target made of titanate lead lanthanum is sputtered by an rf sputtering method to form the PLT film 204 which is oriented to the c-axis on the surface of the MgO monocrystal substrate 221 (FIG. 29 (a)). Next, polyimide resin is applied except for the upper surface of this PLT film 204, and the first layer of polyimide resin film 222 is formed. Then, on top of this layer, the lower lead electrode film 223 of Ni-Cr is formed by a sputtering method (FIG. 29 (b)). Furthermore, polyimide resin is applied on the surface to form the second layer of polyimide resin film 232 (FIG. 29 (c)). The MgO monocrystal substrate 221 disposed thereon with the above-mentioned multilayer film composition is reversed and placed on top of the ceramic substrate 207 made of ceramics such as alumina, which is disposed with a through-hole in the central part having a rectangular cross-section. Then, the MgO monocrystal substrate 221 and the ceramic substrate 207 are adhered and fixed by the adhesive 230 (FIG. 29 (d)). After adhering, for the purpose of improving thermal sensitivity of the PLT film 204, MgO in the MgO monocrystal substrate 221 is etched and removed completely except for the multilayer film composition formed on the MgO monocrystal substrate 221 (FIG. 29 (d)). By removing MgO, the PLT film 204 is newly exposed to the surface, and the Ni-Cr upper lead electrode film 225 is formed thereon. In addition, the connection electrodes 209 and 208 which were formed in advance on top of the ceramic substrate 227 are connected with the upper lead electrode 225 and the lower lead electrode 223 by means of the conductive paste 206 and 216 (FIG. 29 (f)). In this way, the conventional pyroelectric infrared sensor element can be obtained.
Although the semiconductor type acceleration sensor mentioned above can be miniaturized by integration using semiconductor techniques, this acceleration sensor has the problem of having complicated manufacturing steps, since it is necessary to form a dead weight part or a cantilever part by applying micromachining techniques such as anisotropic etching using an alkaline solution. For example, the curved gage type uses an anisotropic etching technique for forming a cantilever part, but it is difficult to control the thickness etc. of the cantilever part. Also, in order to attain shock resistance and resonance resistance, when the cantilever part for the support of a dead weight part comprises a plurality of components, each part was required to be accurate in size etc., so that the manufacturing steps became even more complicated.
As for the conventional pyroelectric infrared sensor element mentioned above with reference to FIG. 29, the PLT film 204 comprising a pyroelectric dielectric oxide film is held only by the polyimide resin films 222 and 223, and the above-noted polyimide resin films 222 and 223 are held at the circumference by the ceramic substrate 207. Therefore, because of the contraction etc. caused by the difference in material characteristics between the polyimide resins 222, 223, the PLT film 204, and the ceramic substrate 227, the lead electrodes 223 and/or 125 were vulnerable to disconnections. A further problem was that the polyimide films 222 and/or 223 which hold the PLT film 204 tended to crack easily.
Furthermore, since the manufacturing method of the conventional pyroelectric infrared sensor element uses an expensive MgO monocrystal substrate which has a cleavage plane of (100) and has a mirror that must be polished, the infrared sensor element also becomes expensive. Moreover, after the pyroelectric dielectric oxide film was formed, the MgO monocrystal substrate 221 which was placed directly below the PLT film 204 needed to be removed carefully by an etching method. Thus, this manufacturing method has the problem of being complicated.